schrieb Marcus D. Leech on 2011-01-13 04:56: > On 01/12/2011 10:46 PM, Brian Padalino wrote:
> I'm increasingly liking the approach where you "demarc" at the digital > output of the ADC that I suggested > earlier where you terminate in something like a LPC-FMC connector or > something equally > convenient, which allows you to adapt to various "getting bits to the > host" approaches. > Including: > > o 1GiGe > o USB-2.0 > o USB-3.0 > o PCIe o direct (embedded) Any chance to transfer via eSATA? Quite common nowadays. > But maybe that's the road to *more* expensive, not less (although cost > is only ONE of the factors > of a project like this). First, price is one constraint, or one of the features, that would make such a device attractive to a wide audience. Others would be flexibility, simpleness, that is fitness for home building, capabilities: lowest/highest frequency, dynamic range, simultaneous bandwidth, DDC, filtering, FPGA; RX only vs.s RX and TX; interface and others. For price: I'd say 200$-300$ for a tunable frontend, 16bit resolution, >500kSps and flexibility to support IF input, clock input, digital stream output (for other transport/consuming system) DIY-kit is attractive. If the flexibility is not available, this would be to expensive. Phillip Balister tried to connect a USRP to a Beagle Board, which was not so easy because the USRP hat no way to tap the data between the FPGA and the USB interface. For flexibility, being able to bypass stages or feed signals e.g. at the ADC would be cheap. Preparing for different transport systems would make it more future proof. If USB3.0 hardware support is not satisfying now, maybe it is in two years. For the data bus part this would require a prepared interface for data and control lines. I see the target for such at above all the soundcard solutions and below the USRP1. USRP1 can do 8MHz Bandwidth complex at 14bit/sample RX, which is more than enough for hobbyists. What would be interesting for university teaching and research? What would be interesting for other potential users, like hams? Did I miss some? USB3.0 is common at new high end PCs and Laptops, ExressCard at Laptops. Of course every PC with PCIe is ubgradable to both of them, but it's extra money to spend and extra hassle to get it started. I do not see urgent need for such extreme data rates. This could be a second step. For a start, a not too expensive, but still capable system with options for extension seems most doable for me. If the transport systems is to be fully integrated in the first shot, at least the data should be accessible via some interface. Patrick -- Engineers motto: cheap, good, fast: choose any two Patrick Strasser <patrick dot strasser at tugraz dot at> Student of Telematik, Techn. University Graz, Austria _______________________________________________ Discuss-gnuradio mailing list Discuss-gnuradio@gnu.org http://lists.gnu.org/mailman/listinfo/discuss-gnuradio